The systematic review was reported according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 Statement and Checklist (Checklist 1) and abstract checklist (Checklist 2) [38]. It was registered at PROSPERO (CRD42022301810).

This study is a systematic review of the literature. Ethical approval was not necessary.

We searched in five electronic databases (PubMed [MEDLINE], Embase, Web of Science [Science Citation Index & Social Sciences Citation Index], PsycInfo, and Epistemonikos Database). To identify ongoing studies, we also searched five clinical trial registries (Cochrane Central Register of Controlled Trials, Cochrane COVID-19 Study Register, ISRCTN registry, Australian New Zealand Clinical Trials Registry, and EU Clinical Trials Register). Additionally, we screened preprint servers (medRxiv), published conference proceedings, and gray literature (Base, opengrey.eu, Social Science Open Access, Centers for Disease Control and Prevention WONDER, and ProQuest). We manually searched the reference lists of included studies and contacted corresponding authors to obtain unpublished study reports. The initial search was performed in May 2022, followed by a first and second update search in December 2023 and August 2025. Only studies published after the onset of COVID-19 (December 2019) were considered; a filter was applied accordingly. The search strategy was not restricted in terms of the population of interest or language. The review question was operationalized based on the PICO (Population-Intervention-Comparator-Outcome) framework, which was extended by the study design [39]. The search strategies are presented in detail in the additional file Multimedia Appendix 1 [40-42].

The results of the study search and selection process are shown in Figure 1. The search yielded a total of 51,171 papers, of which 188 studies were deemed relevant. Among these, 89 studies addressed the target population of non-HPs, while 99 focused on HPs. Given the considerable number of potentially relevant studies, the present systematic review primarily focused on the population of HPs. The results about other populations will be presented in a separate publication. For future updates, we identified ongoing studies for potential inclusion; however, only published studies have been included in the present analysis to report the outcomes. In the additional file Multimedia Appendix 2 [61-188], Table S1 shows the excluded studies with reasons provided.

Of the included studies, there were 15 RCTs [61-74], 4 NRSIs [75-78], 74 uncontrolled before-and-after studies [79-138,188], 5 ongoing studies [139-143], and 1 study as awaiting classification [144]. The references were published between 2020 and 2025. Most studies (96/99 studies, 97%) were published in English and 3 studies were published in Spanish. In Multimedia Appendix 2, the Tables S2-S4 show the summarized characteristics of included studies.

In total, 329,623 HPs were enrolled in interventional studies, ranging from 18 to 234,643 participants per intervention. Most participants were enrolled in uncontrolled before-and-after studies (n=327,298), whereas substantially fewer participants were included in RCTs (n=2034) and NRSIs (n=291). Most interventions were directed toward a broad range of mixed health occupational groups, accounting for 43% of studies (40/93 studies, 298,220 participants). Interventions targeted particularly at students from various disciplines, including medicine, nursing, dentistry, or pharmacy, were examined in 29% of studies (27/93 studies, 4388 participants). Five of the 93 (5%) studies were conducted in both HPs and non-HPs (9068 participants). Table 3 shows the distribution of study populations by study design in more detail. RCTs were predominantly conducted among students and specialist medical practitioners, whereas uncontrolled before-and-after studies more frequently included mixed HP groups, nursing professionals, pharmacists, midwives, and students. Therefore, the distribution of professional groups differed across study designs.

The average age of participants was 29.9 (SD 8.8) years, as reported in 37% (34/93 studies) of studies. Further, 64% of the participants were female (65 studies). Most studies were conducted in Asia (48/93 studies, 52%), North America (17/93 studies, 18%), and Europe (15/93 studies, 16%). Overall, of the 93 included studies 39 (42%) took place in high-income countries, 27 (29%) in lower-middle countries, 26 (28%) in upper-middle-income countries, and 3 (3%) in low-income countries.

The type of intervention, delivery mode, and methods, settings, and comparator varied widely across the included studies in Table S2 in Multimedia Appendix 2. We identified three predominant intervention types: (1) educational interventions, (2) training programs, and (3) blended intervention approaches. Overall, 61 of the 93 (66%) studies applied educational interventions with traditional teaching of theoretical knowledge. Six (6%) studies provided training programs comprising demonstrations and the performance of practical, hands-on skills (eg, donning and doffing of PPE). Within 26 of the 93 (28%) studies, blended interventions were conducted, which combined both theoretical and practical approaches. While educational interventions were the most common intervention type across all study designs, they were predominantly evaluated in uncontrolled studies. Training programs were mainly evaluated in controlled study designs, whereas blended approaches were evaluated in both controlled and uncontrolled studies (Table 4). In controlled studies, comparators most commonly consisted of educational approaches delivered in formats similar to the intervention (Table 5), indicating that many studies compared alternative educational strategies rather than using a passive control group.

HPs (27/93 studies, 29%), health educators (10/93 studies, 11%), trained instructors (13/93 studies, 14%), and researchers (18/93 studies, 19%) delivered the interventions. Universities (23/93 studies, 25%) and hospitals (33/93 studies, 35%) were the most prevalent settings. The duration of interventions lasted from 10 minutes to 19 weeks. Most of the interventions were provided in groups (50/93 studies, 54%), followed by one-to-one sessions (25/93 studies, 27%). Online distance learning (43/93 studies, 46%) was more prevalent than in-person teaching (29/93 studies, 31%) and hybrid learning (11/93 studies, 12%).

Among 93 studies, 41 (44%) reported a theoretical framework or model for guiding processes such as development, implementation, and evaluation of the intervention. Hereby, 6 studies (6%) referred to learning models (Peyton’s approach [64], Miller’s pyramid-based training [127], cognitive learning theory [91,100,103], and simulation-based mastery learning [116]) and 2 (2%) applied health behavior change models (eg, health belief model [66] and theory of change model [112]). Further, two (2%) studies were based on Kirkpatrick’s evaluation model [110,145]. Health manuals, COVID-19 protocols, and clinical information (eg, WHO protocols and Infection Prevention Control guidelines) were applied in 27 (29%) studies. While 2 (2%) of the included studies incorporated HL as a component within the curriculum [188] or as an underlying concept for an online learning platform [121], no study explicitly referred to an HL model or provided an established definition.

The vast majority of studies reported primary outcomes related to changes in COVID-19 knowledge (75/93 studies, 81%). Of the 93 studies, 20 (22%) assessed general knowledge of COVID-19, while 43 (46%) focused on specific knowledge outcomes, primarily concerning infection prevention and control measures (19 studies, 20%), PPE performance (8 studies, 9%), COVID-19 vaccine (9 studies, 10%), SARS-CoV-2 transmission (5 studies, 5%), and infection management (4 studies, 4%). In 21 (23%) studies [62-64,67-69,72,73,75,78,84,91,92,97,106,116,122,127,129,146,147] changes of COVID-19 infection prevention skills were targeted, with enhancement of PPE performance being the most frequently addressed outcome (13 studies, 14%).

In total, 38 of the 93 (41%) studies [62,65,66,74,76,78-80,83,84,86,87,90,93,97,101,104,105,108,111,112,119,123,130-132,135,146-155,188] reported secondary outcomes such as confidence and self-efficacy (19 studies, 20%), compliance toward and prevalence of infection prevention measures behavior (6 studies, 7%), attitudes (11 studies, 12%), and composite outcomes (21 studies, 23%). One study [121] developed an online learning platform aimed at fostering HL and focused in particular on knowledge, attitudes, and practices. Another study [62] reported adverse events related to motion sickness caused by the use of virtual reality simulation. Among 21 studies (23%) a discrepancy was observed between the reported outcomes and the items actually used to measure the intended construct.

Most studies (65/93 studies, 70%) measured outcomes at short-term follow-up, defined as immediately following the intervention for up to four weeks from the program’s start. Of 93 studies, 8 (9%) reported medium-term follow-up, ranging up to six months. Additionally, 9 (10%) studies determined outcomes at a second follow-up ranging from one week to eight months after completing the intervention. In 19 (20%) studies, the time point of the postintervention outcome assessment was not clearly reported.

The assessment differed across all 93 studies. The content of the assessment tools was based on hospital internal protocols [68,72,73,83], national and international guidelines, such as those provided by the WHO [84,91,100,146,156], the Centers for Disease Control and Prevention [91,116,134], or the Ministry of Health [84,153]. Further, 60 studies (65%) applied self-developed instruments, created for the study. Additionally, 26 (28%) studies reported information on the development of the assessment tool. Among 37 (40%) studies using a validated tool for outcome measurement, 11 (12%) studies did not provide details regarding the validation process and measurement properties. Most studies (82/93 studies, 88%) used participants’ reported measures with Likert scales, multiple-choice, “true”/“false” or “yes”/“no” response options. In 17 of the 93 (18%) studies [62-64,69,73,75,83,84,90-92,97,106,116,122,127,129], outcomes were assessed through performance-based evaluations conducted by an observer (eg, teacher, instructor, or member of the study team) using a checklist. In one study [188], the assessment tool included a HL-related item specifically.

Overall, the included studies indicated substantial conceptual overlap between COVID-19–related HL and educational or skills-based infection control interventions. Most studies focus on knowledge and procedural competencies, corresponding to the “understand” and “apply” facets, rather than explicitly addressing the full construct. By contrast, no study explicitly assessed COVID-19–related HL as an overall construct or examined changes in “access” and “critical appraisal” of health information (Table 6).

An overview of the risk of bias assessments of included studies is presented in detail in the additional file Multimedia Appendix 3 [61-138,145-158,188]. Among primary outcomes of RCTs (k=15) assessed with the Cochrane RoB 2 Tool, 14 of 26 (54%) compared results of primary outcomes were evaluated as “some concerns.” Primary outcomes of 6 of 26 (23%) results had “low” risk of bias, while another 6 of 26 (23%) results were rated at “high” risk of bias. Most results (16/26 studies, 62%) were rated at high risk of bias regarding the selection of the reported results due to a lack of prespecified statistical analysis plans or study protocols. Some studies [61,63] did not adequately handle missing outcome data.

Based on the ROBINS-I assessments for NRSIs, all six evaluated outcomes of 4 included studies were judged as having “serious” risk of bias. Pertaining to the risk of bias of uncontrolled before-and-after studies using EPHPP-QAT, 73 of 74 (99%) studies were rated as being “weak,” and 1 study as “moderate.” Most uncontrolled studies used a nonprobability purposive or convenience sampling technique, resulting in a higher risk of selection bias. Neither did they report information on the blinding of participants nor outcome assessors. All NRSI and uncontrolled before-and-after studies are prone to confounding (eg, preexisting COVID-19 knowledge, interest in gaining more knowledge, and socioeconomic status), but no study adjusted or controlled for potential confounding.

We performed meta-analyses for changes of COVID-19–related knowledge and skills. Hereby, we restricted meta-analyses to RCTs and primary outcomes due to high heterogeneity across interventions and outcomes in both NRSIs and uncontrolled before-and-after studies and due to inconsistencies in secondary outcome definitions in RCTs. The results are presented in the summary of findings table, along with GRADE certainty of evidence ratings (Table 7). The summary of findings tables for all included studies, separated by study design, are shown in the additional file Multimedia Appendix 4 [61-138,145-158,188,189].

Data from four RCTs were considered for the meta-analysis on the outcome COVID-19–related knowledge, grouped into knowledge on vaccine and PPE (Figure 2). COVID-19-relevant HL interventions may increase vaccine knowledge at postintervention (standardized mean difference [SMD] 1.00; 95% CI 0.33 to 1.67, I²=24%), but the evidence is very uncertain about the effect. Further, the evidence is uncertain and suggests that interventions may not increase PPE knowledge at postintervention (SMD 0.21; 95% CI −0.09 to 0.51, I²=0%).

Of the 15 RCTs included, 7 RCTs contributed data to the meta-analysis (Figure 3). They reported on COVID-19–related infection prevention and control performance and were categorized into competencies in PPE donning, doffing, or both. The HL-relevant interventions may increase postintervention donning and doffing skills, though the evidence remains very uncertain (SMD 1.95; 95% CI 0.82 to 3.09, I²=46.1%). The evidence is very uncertain about the effect of interventions on doffing skills at postintervention (SMD 0.42; 95% CI −0.95 to 1.78, I²=62.8%) and at follow-up (SMD 0.61; 95% CI −0.07 to 1.29, I²=0%), as well as on donning skills at postintervention (SMD 0.28; 95% CI −0.48 to 1.04, I²=0%).

In addition to the findings of the meta-analysis, a narrative description of the 74 uncontrolled before-and-after studies is provided in Multimedia Appendix 5 [94,107,112].

This systematic review aimed to identify and critically evaluate evidence pertaining to interventions that focus on enhancing COVID-19–related HL in HPs. Meta-analysis revealed favorable changes in COVID-19–related infection prevention and control skills after the intervention, particularly in the performance of PPE donning and doffing, which are essential competencies for reducing the risk of health care–associated infections. These findings are similar to those of Saunders et al [190], who systematically investigated the impact of HL training in HP students.

However, the pooled effect sizes must be interpreted with caution. Although some effect sizes were large, they were derived from a small number of studies with limited sample sizes, which increases the risk of imprecision and unstable effect estimates. Small studies are also more susceptible to bias and may overestimate intervention effects. This is reflected in the GRADE assessment, which indicated low to very low certainty of evidence due to risk of bias, inconsistency, and imprecision. Further, many interventions targeted only a subgroup of the population of interest, for example, medical students. Whether the reported results can be applied to the population of HPs at large remains unclear [9]. Thus, the identified body of evidence can largely be characterized by high inconsistency. While the pooled effect estimate suggested a statistically significant effect for the outcomes COVID-19 vaccine-related knowledge and competencies in PPE donning and doffing at postintervention, the PI included nonsignificant and both potentially beneficial and harmful effects, indicating considerable variability and limiting the generalizability of effects across intervention approaches, HP occupational populations, settings, and outcomes. This finding is consistent with the results of similar reviews [190,191].

Moreover, the small sample sizes in most of the included studies may have led to a failure to detect meaningful differences due to insufficient statistical power. They might be attributed to the high workload of HPs, which was exacerbated during the onset of the pandemic, and associated with time constraints, limited personnel resources, and high rates of illness among infected staff, perceived exhaustion, and pandemic fatigue [6,8]. Many studies used nonprobability purposive or convenience sampling techniques, which are associated with a higher risk of bias. This might also be linked to the ad hoc nature of the interventions, which were implemented in response to the pandemic. Further, the high risk of bias in most studies weakens confidence in the findings. NRSIs are prone to confounding. For instance, participants may already possess a higher level of knowledge about COVID-19 due to extensive media coverage or a general interest in acquiring more information. In uncontrolled before-and-after studies, we cannot with certainty attribute observed effects to the intervention, as there is no control group.

None of the included studies were based on a theoretical framework of HL, aligning with findings of previous studies [7,37,191]. It appears that COVID-19 HL has more often been addressed in cross-sectional than in interventional studies [7,37]. Baumeister et al [191] concluded that the integrated HL model [20] might not be as familiar with the model as desired among researchers. More than half of the studies were conducted in Asia and North America, regions where the model might be less known, as it was developed in Europe [191,192].

Further, the identified body of evidence revealed conceptual overlap between COVID-19–related HL and educational or skills-based interventions in the context of COVID-19 prevention and infection control. Most included studies were not designed as explicit HL interventions. Instead, COVID-19–related HL was frequently operationalized indirectly through knowledge- and skills-related indicators rather than validated HL instruments. These indicators were relevant to the present review as they could be mapped onto specific HL facets, particularly “understand” and “apply.” However, they should not be interpreted as equivalent to the full construct of COVID-19–related HL. None of the studies examined all four facets of HL, and the facets relating to “access” and “critical appraisal” were not explicitly addressed. The available evidence should therefore be understood as HL-relevant, but only partly HL-specific. Nevertheless, the theoretical foundation remains essential for systematically developing, comparing, and refining interventions [193]. Most of the studies investigated socio-cognitive outcomes combined, such as knowledge, attitudes, and practical skill acquisition, which align more closely with the Knowledge, Attitudes, and Practices framework [194].

It should also be noted that the intended outcomes were primarily assessed subjectively through self-report. Self-assessments of participants pertained to self-reported judgments about how well a person perceives themselves to know COVID-19–specific health information. This often leads to over- or underestimation of their own knowledge or confidence, and can also evoke social desirability bias, which might affect confidence in the results [195]. Miller [196] highlighted the significant impact of measurement moderators on the effect sizes of HL interventions. Their findings indicated that effect sizes, and consequently the observed improvements, were greater when HL was assessed subjectively rather than through objective measures by performance or achievement indices [196].

Moreover, the constructs used in the instruments were often merged, making it unclear which specific outcome was being measured, potentially leading to over- or underestimation of the results within the study. In 21 studies, a mismatch between the reported intervention content and the measured outcomes was observed. This finding is also consistent with the results of Walters et al [193]. Therefore, uncertainty regarding composite outcomes cannot be ruled out.

In more than half of the studies (60 studies), new questionnaires were developed; however, 26 studies reported the development process. Limited information regarding the development process and the origin of the items was observed. These findings compare with those of Grepmeier et al [7]. The lack of detailed information could be attributed to the need for implementing and conducting interventions rapidly, given the urgency to quickly develop and foster pandemic-related competencies and knowledge. Instruments for assessing COVID-19–related HL in respective countries may not yet be widely available or accessible in various languages. Currently, only a few instruments address COVID-19 HL, such as the validated tools developed in Europe for the general population [17,197] and specifically for HPs, such as the “COVID-19 related health literacy in health professionals” [43].

Regarding the delivery mode, most of the interventions were conducted through online distance learning, likely due to the circumstances and contact restrictions during the pandemic. Nevertheless, this demonstrates how quickly the implementation of interventions was adapted to rapidly changing external conditions.

Although the results indicate that COVID-19 HL interventions are beneficial for selected outcomes, individuals with high levels of self-reported and observable skills may still encounter real-world challenges in applying these competencies. This can occur, for example, in unfamiliar environments or when interacting with individuals they find intimidating, such as those in hierarchically higher-ranking positions [35]. Whether HPs are able to apply what they have learned over a longer period and translate it into actual behavior cannot be confirmed based on the available data. Consistent with similar studies [190,191], the outcomes were primarily measured immediately or shortly after the intervention was conducted. No long-term follow-ups, extending up to one year, were reported. This may be due to the urgency and relevance of rapidly providing interventional effects to the scientific community. As a result, it remains unclear to what extent COVID-19–related HL interventions have long-term sustainability.

This review has several limitations. Substantial heterogeneity in interventions, comparators, settings, HP occupations, and outcomes limited the comparability between studies. Therefore, we were not able to state the effect of the interventions separated by delivery mode (eg, in-person and online distance learning), and delivery methods (eg, interactive and self-directed), intervention types, or specific occupational HP groups (eg, nurses, physicians, and medical students). Given the limited comparability across studies and substantial heterogeneity, the meta-analyses were conducted in an exploratory manner. Accordingly, the observed effects should be interpreted in light of the high risk of bias and low to very low certainty of evidence according to GRADE, and therefore interpreted with caution and regarded as preliminary evidence rather than definitive conclusions.

Moreover, for certain outcomes, a meta-analysis was not considered appropriate due to substantial differences in outcome metrics across studies. Consequently, the GRADE assessment could only be applied when meta-analytical results were available, and the domain of inconsistency could not be evaluated for these outcomes. The findings for these outcomes were reported narratively. We encountered challenges in contextualizing the results and the effect sizes within the existing literature, as there are currently only a few comparable reviews that report a meta-analysis based on HL interventions [190,191,198,199].

Further, most included studies did not explicitly aim to enhance the full construct of COVID-19–related HL. Instead, most studies assessed knowledge- and skills-based outcomes that were interpreted in this review as HL-related indicators. As a result, the review captures an evidence base that is closely related to HL, but not consistently grounded in explicit HL theory or measurement. In terms of the meta-analysis, it was not possible to determine publication bias due to the small number of studies included (≤10 studies). Additionally, a considerable proportion of effect sizes required conversion, and in certain cases, the assumptions underlying conversion could only be partly verified due to limited data availability in the primary studies.

A further limitation is the use of Microsoft Excel for data extraction. Although this program is commonly used for extracting data, it lacks the automated error-checking and audit functions of specialized systematic review software. To minimize potential error, data extraction was performed by two reviewers; one reviewer extracted relevant data, and the second reviewer checked the data extraction table for accuracy.

While these limitations should be considered, the present review also offers notable strengths. To our knowledge, this is the first study systematically synthesizing and summarizing evidence on interventions aiming to increase COVID-19–related HL in HPs. We ensured a comprehensive synthesis approach by searching 15 databases and registries, and including NRSIs and uncontrolled before-and-after studies in addition to RCT evidence, while closely adhering to guidelines for systematic reviews [42]. We followed the integrated model by Sørensen et al [20] as a framework for grouping and interpreting outcomes and intervention approaches. Further, we used validated assessment tools for rating risk of bias [47-49] and conducting synthesis in light of the certainty of the evidence by using GRADE [50]. The interpretation and reporting of the meta-analytic findings throughout this review were aligned with the corresponding GRADE assessments.

Our findings imply a need for further measures to be taken. Future intervention studies should be based on established HL concepts as a theoretical framework for designing and implementing interventions to promote COVID-19–related HL as an essential key competency in HPs on the individual level. Existing models, such as the conceptualization of HL by Sørensen et al [20], the Medical Research Council framework [200], or the Optimizing HL and Access approach [201], offer valuable foundations for the systematic development and adaptation of HL interventions. Further, as Nutbeam and Lloyd [35] concluded, a consolidated standard curriculum for enhancing HL among HPs would be essential to adapt and tailor methods in response to crises such as the pandemic. A toolbox of established targeted educational and practice-based approaches for communicable diseases might guide researchers and clinicians in developing and implementing HL interventions rapidly in acute situations. For instance, the facet “understand” was frequently assessed through performance tests or self-reported questionnaires in order to capture participants’ COVID-19 knowledge. A more effective method for assessing whether participants have understood the content of a training program might be using the widely applied communication technique “teach-back,” where participants repeat in their own words what they have understood [35]. The “apply” facet may be better evaluated through observations conducted by a trained reviewer, using a predeveloped checklist (reported in 17 studies) compared to the often used self-assessments.

Moreover, the use of validated instruments available in various languages to evaluate changes in COVID-19–related HL is emphasized. Further, to monitor changes over time and capture long-term effects, future trials should incorporate repeated follow-up assessments after the intervention for at least 12 months after the intervention.

This review focused on the individual HL of HPs. The representation of professional HL in the context of the pandemic was not within the scope of this work. Beyond individual skills, clinicians with higher COVID-19–related HL, and thus a deeper understanding of COVID-19 and its related guidelines, may be better positioned to make informed decisions regarding patient management. This includes recognizing symptoms, knowing when to increase care, advising on appropriate treatments, answering patient questions effectively, and providing clear instructions on infection prevention and control measures. This is especially critical in environments in which misinformation and frequently changing guidelines can create uncertainty for both providers and patients [30,202]. Moreover, HPs are among others the primary sources of information for patients and the public [30,203]. Therefore, the level of COVID-19–related HL among HPs may also influence their surroundings, including the behaviors and attitudes of patients.

In this context, professional HL plays a crucial role, as it extends beyond individual competencies. Professional HL refers to the ability to communicate health information to patients in an understandable manner and to act on this information [204]. Pandemic-related HL interventions should not only focus on enhancing individual skills but also address professional HL, as HPs play a key role in promoting HL within the general population. Their transferable skills can significantly contribute to improving public HL [35,204].

HL-targeted interventions may empower individuals to enhance their competencies regarding infection control measures, enabling them to navigate the vast amount of pandemic-related information more effectively. This is particularly important as the pandemic transitions into an endemic phase, where ongoing community transmission necessitates sustained health education efforts. Given the ongoing presence of SARS-CoV-2 and the anticipated infection rates due to seasonal factors [205], we assume that the evidence landscape from COVID-19–related HL interventions continues to evolve. The increased number of records identified through database searches between the initial and updated searches (Figure 1) supports this assumption. Therefore, we aim to continuously identify and integrate new evidence and thus transform this systematic review into a living systematic review.

A total of 15 studies investigated the effect of interventions relevant to COVID-19–related HL among HPs. However, the identified evidence base was heterogeneous in terms of intervention content, delivery, study populations, and outcomes, limiting the comparability and the synthesis of effects. Most included studies evaluated short-term educational or skills-based infection control interventions and operationalized HL indirectly through knowledge- and skills-related indicators rather than explicit, theory-based HL measures.

Overall, the identified interventions might support HPs’ short-term competencies relevant to infection prevention and control measures. Although some interventions showed large effects on specific outcomes, the evidence remains uncertain due to small sample sizes, high risk of bias, heterogeneity, imprecision, and limited theoretical grounding. In the absence of long-term follow-up data, the sustainability of observed effects remains unclear. The available evidence is closely related to HL, but not fully representative of explicitly HL-based interventions. Interventions explicitly designed to enhance the full construct of COVID-19–related HL remain largely absent. However, high-quality RCTs with adequate statistical power are needed to advance the current understanding of how interventions can strengthen HL in the context of communicable diseases, including COVID-19.